Signals of Interest Recovery With Multiple Receivers Using Reference-Based Successive Interference Cancellation for Signal Collection Applications

In this paper, we introduce a novel but intuitive scheme to recover multiple signals of interest (SoI) from multiple emitters in signal collection applications such as signal intelligence, electronic intelligence, and communications intelligence. We consider a case where the SoIs form a heavy interference environment. The scheme, which is referred to as reference-based successive interference cancellation (RSIC), involves a combination of strategic receiver placement and signal processing techniques. The scheme works by placing a network of cooperative receivers where each receiver catches its own SoI (despite multiple interferences). The first receiver demodulates the initial SoI (called a reference signal) and forwards it to the second receiver. The second receiver collects a received signal containing the second SoI but is interfered with by the initial SoI, which is a problem called co-channel interference in cellular communications. Unfortunately, the amplitude scaling of the interference is unknown in the second receiver and therefore has to be estimated via least squares error. It turns out that the estimation requires a priori knowledge of the second SoI, which is the very signal it tries to demodulate, thereby yielding a Catch-22 problem. We propose using an initial guess on the second SoI to form an amplitude estimate such that the interference is subtracted (cancelled) from the collected measurement at the second receiver. The procedure is applied to a third receiver (or multiple receivers) until the last of the desired SoI is separated from all of the co-channel interferences. The RSIC scheme performs well. Using quaternary phase shift keying as example modulation, we present major symbol error rate (SER) performance improvements with the use of RSIC over the highly degraded SER of receivers that are heavily interfered and do not employ any cancellation technique.